AI-Exposed Majors Are Getting Hit Hardest by the Job Market

The degree that was supposed to be a golden ticket is starting to look like a lottery ticket. New research from the U.S. Census Bureau finds that graduates from AI-exposed majors saw their odds of landing a job right after graduation fall by five percentage points after ChatGPT's arrival, while their initial full-quarter earnings dropped 13 percent — an earnings decline the researchers describe as "comparable in magnitude to the earnings losses associated with graduating into a large recession," according to The Register's coverage of the working paper.

The paper, published in September 2026 by three Census Bureau economists, is the heaviest statistical artillery yet behind the growing sense that AI is eating the entry level first. It analyzed administrative records covering roughly 29 percent of U.S. bachelor's degrees awarded between 2016 and 2024, comparing outcomes across majors before and after the generative AI boom. That list reads like a roll call of the hardest-hit AI-exposed majors: computer science, computer and information systems, and computer engineering saw the largest declines in both employment and earnings, alongside mathematics, statistics, some engineering disciplines, networking — and, the authors note, journalism.

Where did the earnings go? About half of the decline came from graduates drifting into lower-paying industries — retail counters and restaurant floors — rather than the software and IT jobs those degrees were built to feed, the researchers found. The other half came from lower pay inside the industries that did hire them.

The timeline: from chatbot launch to recession-scale fallout

The damage tracks the technology with unsettling precision. ChatGPT debuted in late 2022. A second Census analysis found that over the ten quarters that followed, employment for workers ages 22 to 24 in the most AI-exposed industry regions fell roughly one-eighth, driven mostly by fewer hires rather than layoffs — people already on payroll were not the main story, according to the Daily Caller's survey of the research.

The anxiety shows up everywhere in the numbers. The unemployment rate for recent college graduates sat near 5.6 percent in the second quarter of 2026, with underemployment at 42 percent, according to the New York Fed. Americans put their odds of finding another job after losing one at just 45.4 percent in August, and the share expecting unemployment to rise hit its highest mark since the depths of the COVID pandemic. Handshake's data on the Class of 2026 shows job postings down 12 percent from pre-pandemic levels, while 62 percent of graduating seniors describe themselves as pessimistic about the job market.

Corporate America is not helping the mood. Employers attributed more than 116,000 planned job cuts to AI through August — more than one-fifth of all announced cuts this year, per Challenger, Gray & Christmas. Oracle raised the expected cost of its fiscal 2026 restructuring by hundreds of millions of dollars while ramping AI spending, and Microsoft cut roughly 500 Xbox jobs in September after a far larger round in July, according to the Daily Caller's reporting on corporate announcements.

Why the entry level absorbs the shock first

Economists call the current environment a low-hire, low-fire labor market: workers who already have jobs rarely lose them, but anyone trying to get in faces a wall. Brad Hershbein, a senior economist at the W.E. Upjohn Institute for Employment Research, told the Daily Caller that this hurts young workers most because missing a strong first job can shape earnings and career progression for years. If AI compresses the bottom rungs of the ladder, graduates from AI-exposed majors lose the very on-ramp that was supposed to justify the tuition.

The hollowing reaches beyond degree-holders. At a recent Senate subcommittee hearing on AI and the workforce, Justin Heck of Opportunity@Work warned lawmakers about "gateway jobs" — the customer-service and entry-level roles that have long served as stepping stones into careers. About half of AI-exposed occupations are gateway roles, Heck noted, according to Governing's workforce analysis. The danger is not mass layoffs but a ladder that quietly stops rising: nobody gets fired, but nobody gets promoted either, and the on-the-job learning that once built careers evaporates.

The caveats the researchers insist you hear

The Census economists themselves urge caution before anyone declares the career script dead. They acknowledge several alternative explanations — and found that alternative paths like graduate school and self-employment could account for as much as half of the employment decline. They also told The Register that some firms may be hiring fewer graduates simply because AI makes the future feel uncertain, not because the technology itself replaced those roles.

There is also a crucial asymmetry in the data. Graduates from less AI-exposed fields — including nursing and many education majors — have not experienced the same deterioration in early-career employment, which is exactly the opposite of a normal recession. And the pain fades with time: the effects diminish as graduates move further from labor-market entry, though research on Great Recession cohorts suggests that rough starts can leave workers trailing on lifetime earnings. The authors caution that they are "very cautious in speculating" about AI's long-term impact on the whole economy from short-term effects on AI-exposed majors.

What should new graduates actually do? The one lever the data supports is skills. In Dexian's 2026 Work Futures research, 72 percent of workers said improving their skills could lead to higher pay, advancement, or more interesting work. Practically, that means targeting less-exposed lanes, building AI fluency rather than competing with it, and treating internships and apprenticeships — the gateway jobs that still exist — as the real first interview. The recession comparison in the Census paper is about magnitude, not cause: this downturn has a narrow target — AI-exposed majors —, and knowing where it points is the first step to dodging it. For more deep data reads on the forces reshaping work, see Deep Dives, and for how workers are redesigning the workweek itself, read Microshifting: Why Workers Are Ditching the 9-to-5.